Your SlideShare is downloading. ×
Screening and Prostate-Cancer Mortality  in a Randomized European Study
Upcoming SlideShare
Loading in...5
×

Thanks for flagging this SlideShare!

Oops! An error has occurred.

×

Saving this for later?

Get the SlideShare app to save on your phone or tablet. Read anywhere, anytime - even offline.

Text the download link to your phone

Standard text messaging rates apply

Screening and Prostate-Cancer Mortality in a Randomized European Study

701
views

Published on

Published in: Health & Medicine, Business

0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total Views
701
On Slideshare
0
From Embeds
0
Number of Embeds
0
Actions
Shares
0
Downloads
12
Comments
0
Likes
0
Embeds 0
No embeds

Report content
Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

Cancel
No notes for slide

Transcript

  • 1. The n e w e ng l a n d j o u r na l of m e dic i n e original article Screening and Prostate-Cancer Mortality in a Randomized European Study Fritz H. Schröder, M.D., Jonas Hugosson, M.D., Monique J. Roobol, Ph.D., Teuvo L.J. Tammela, M.D., Stefano Ciatto, M.D., Vera Nelen, M.D., Maciej Kwiatkowski, M.D., Marcos Lujan, M.D., Hans Lilja, M.D., Marco Zappa, Ph.D., Louis J. Denis, M.D., Franz Recker, M.D., Antonio Berenguer, M.D., Liisa Määttänen, Ph.D., Chris H. Bangma, M.D., Gunnar Aus, M.D., Arnauld Villers, M.D., Xavier Rebillard, M.D., Theodorus van der Kwast, M.D., Bert G. Blijenberg, Ph.D., Sue M. Moss, Ph.D., Harry J. de Koning, M.D., and Anssi Auvinen, M.D., for the ERSPC Investigators* A bs t r ac t BackgroundThe authors’ affiliations are listed in the The European Randomized Study of Screening for Prostate Cancer was initiated inAppendix. Address reprint requests to the early 1990s to evaluate the effect of screening with prostate-specific–antigenDr. Schröder at the Erasmus Medical Cen-ter, P.O. Box 2040, Rotterdam 3000 CA, (PSA) testing on death rates from prostate cancer.the Netherlands, or at secr.schroder@erasmusmc.nl. Methods*Members of the European Randomized We identified 182,000 men between the ages of 50 and 74 years through registries Study of Screening for Prostate Cancer in seven European countries for inclusion in our study. The men were randomly (ERSPC) are listed in the Appendix. assigned to a group that was offered PSA screening at an average of once every 4 yearsThis article (10.1056/NEJMoa0810084) was or to a control group that did not receive such screening. The predefined core agepublished at NEJM.org on March 18, 2009. group for this study included 162,243 men between the ages of 55 and 69 years. The primary outcome was the rate of death from prostate cancer. Mortality follow-upN Engl J Med 2009;360:1320-8.Copyright © 2009 Massachusetts Medical Society. was identical for the two study groups and ended on December 31, 2006. Results In the screening group, 82% of men accepted at least one offer of screening. During a median follow-up of 9 years, the cumulative incidence of prostate cancer was 8.2% in the screening group and 4.8% in the control group. The rate ratio for death from prostate cancer in the screening group, as compared with the control group, was 0.80 (95% confidence interval [CI], 0.65 to 0.98; adjusted P = 0.04). The absolute risk difference was 0.71 death per 1000 men. This means that 1410 men would need to be screened and 48 additional cases of prostate cancer would need to be treated to prevent one death from prostate cancer. The analysis of men who were actually screened during the first round (excluding subjects with noncompliance) provided a rate ratio for death from prostate cancer of 0.73 (95% CI, 0.56 to 0.90). Conclusions PSA-based screening reduced the rate of death from prostate cancer by 20% but was associated with a high risk of overdiagnosis. (Current Controlled Trials number, ISRCTN49127736.)1320 n engl j med 360;13 nejm.org march 26, 2009 The New England Journal of Medicine Downloaded from nejm.org on October 8, 2011. For personal use only. No other uses without permission. Copyright © 2009 Massachusetts Medical Society. All rights reserved.
  • 2. European Trial of Prostate-Cancer ScreeningM easurement of serum prostate- were unable to provide the necessary data, and specific antigen (PSA), a biomarker for investigators in France decided to participate in prostate cancer,1 is useful for the detec- 2001, so data from their analyses were not in-tion of early prostate cancer.2 Nevertheless, the cluded because of the short duration of follow-effect of PSA-based screening on prostate-cancer up. Men in whom prostate cancer had been diag-mortality remains unclear.3 The European Ran- nosed (according to data from questionnaires ordomized Study of Screening for Prostate Cancer registries) were ineligible. Within each country,(ERSPC) was initiated in the early 1990s to deter- men were assigned to either the screening groupmine whether a reduction of 25% in prostate- or the control group, without the use of blocks ofcancer mortality could be achieved by PSA-based numbers or stratification on the basis of random-screening.4 Preliminary data from this study have number generators (Fig. 1).been published and can be accessed at www. At all study centers, the core age group includ-erspc.org. Another randomized screening trial in ed men between the ages of 55 and 69 years atthe United States, the Prostate, Lung, Colon, and entry. In addition, in Sweden, study investigatorsOvarian (PLCO) Cancer Screening Trial, was ini- included men between the ages of 50 and 54tiated around the same time, and interim results years, and investigators in the Netherlands, Italy,are also reported in this issue of the Journal.5 Belgium, and Spain included men up to the age of 74 years at entry. In Switzerland, men be- Me thods tween the ages of 55 and 69 years were included, with screening up to the age of 75 years. In Fin-Study Design land, men were recruited at the ages of 55, 59, 63,We designed the ERSPC as a randomized, multi- and 67 years and were screened until the age ofcenter trial of screening for prostate cancer, with 71 years. Screening was discontinued in all otherthe rate of death from prostate cancer as the pri- centers when the chosen upper age limit wasmary outcome. An independent data and safety reached. The validity of randomization was de-monitoring committee reviewed the trial, and termined by comparing the age distributions andinterim analyses were carried out according to a the rates of death from any cause in the two studymonitoring and evaluation plan in which the out- groups.come of the trial was to be presented to the re- At centers in all countries except Finland, sub-search group once a statistically significant re- jects were randomly assigned in a 1:1 ratio to thesult corrected for interim analyses was reached.6,7The study’s protocol was reviewed by local andgovernmental ethics committees (for details, see 182,160 Subjects 50–74 yr old underwent randomizationSupplementary Appendix 4, available with the full 162,387 Were in the core age group (55–69 yr old)text of this article at NEJM.org). Recruitment and randomization procedures 160 Subjects 50–74 yr old died 144 Were 55–69 yr olddiffered among countries and were developed inaccordance with national regulations. In Finland,Sweden, and Italy, the trial subjects were identi-fied from population registries and underwentrandomization before written informed consent 82,816 Were assigned to the 99,184 Were assigned to the screening group control groupwas provided (population-based effectiveness 72,890 Were 55–69 yr old 89,353 Were 55–69 yr oldtrial). In the Netherlands, Belgium, Switzerland,and Spain, the target population was also identi-fied from population lists, but when the men 6830 Had prostate cancer 4781 Had prostate cancerwere invited to participate in the trial, only those 5990 Were 55–69 yr old 4307 Were 55–69 yr oldwho provided consent underwent randomization(efficacy trial). The results of analyses from two Figure 1. Enrollment and Outcomes, According to Age Group at Randomization.participating countries were not included in this The predefined core age groupSchroder study included 162,243 1st be- AUTHOR: for this RETAKE men ICManalysis: investigators in Portugal discontinued tween the ages ofF55 FIGURE: years. REG and 69 1 of 2 2nd 3rdtheir participation in October 2000 because they CASE Revised EMail Line 4-C SIZE ARTIST: ts H/T H/T Enon 22p3 Combo AUTHOR, PLEASE NOTE: n engl j med 360;13 nejm.org march 26, 2009 has been redrawn and type has been reset. Figure 1321 Please check carefully. The New England Journal of Medicine Downloaded from nejm.org on October 8, 2011. For personal use only. No other uses without permission. JOB: 36013 ISSUE: 03-26-09 Copyright © 2009 Massachusetts Medical Society. All rights reserved.
  • 3. The n e w e ng l a n d j o u r na l of m e dic i n e screening group or the control group. In Finland, was used for screening; in 1997, this combination the size of the screening group was fixed at was replaced by PSA testing only.7,11,12 In Belgium, 32,000 subjects. Because the whole birth cohort where the results of a pilot study (from 1991 to underwent randomization, this led to a ratio, for 1994) were included in the final data set up to the screening group to the control group, of ap- 1995, a PSA cutoff value of 10.0 ng per milliliter proximately 1:1.5. was used initially. Most centers used sextant bi- Each center reported data on recruitment, opsies guided by transrectal ultrasonography. As of screening, and mortality twice a year to a central June 1996, lateralized sextant biopsies were rec- data center. Several task forces and working ommended.13 In Italy, transperineal sextant bi- groups were responsible for quality assurance, opsies were used. In Finland, a biopsy procedure including an epidemiology committee, a quality- with 10 to 12 biopsy cores was adopted in 2002 as control committee, a pathology committee, and a general policy for the two study groups. a PSA committee.7 The data and safety monitor- The screening interval at six of the seven cen- ing committee had oversight of the trial, with ters was 4 years (accounting for 87% of the sub- a mandate to stop the trial on demonstrating a jects); Sweden used a 2-year interval. In Belgium, significant difference between the groups or ad- the interval between the first and second rounds verse effects of screening. The monitoring com- of screening was 7 years because of an interrup- mittee received reports on the progress of the tion in funding. trial, including prostate-cancer mortality. Causes of death, which were obtained from registries and Pathological Evaluation individual chart review, were assigned according The primary evaluation of specimens from biop- to definitions and procedures developed for the sies and radical prostatectomies was performed by trial. A committee that analyzed causes of death local pathologists. Central review of the pathologi- was formed at each center, and an international cal analyses was not carried out. However, stan- committee coordinated the work of these na- dardization of procedures was coordinated and tional committees.8,9 achieved by the work of the international pathol- ogy committee. (For details on the committee and Screening Tests and Indications for Biopsy its functions, see Supplementary Appendix 3.) Total PSA was measured with the use of Hybri- tech assay systems (Beckman Coulter). From 1994 Treatment Policies through 2000, the Tandem E assay was used, and The treatment of prostate cancer was performed thereafter the Access assay, with the original according to local policies and guidelines. The Hybritech calibration always applied.10 equality of distribution of treatments that were Most centers used a PSA cutoff value of 3.0 ng applied to the screening group and the control per milliliter as an indication for biopsy. In Fin- group has been evaluated, with little indication land, a PSA value of 4.0 ng per milliliter or more of differences between the two study groups after was defined as positive and the men were referred adjustment for disease stage, tumor grade, and for biopsy; those with a value of 3.0 to 3.9 ng per age (data not shown).14 milliliter underwent an ancillary test — digital rectal examination until 1998 and calculation of Follow-up the ratio of the free PSA value to the total PSA Follow-up for mortality analyses began at ran- value (with a value of ≤0.16) starting in 1999 — domization and ended at death, emigration, or a and were referred for biopsy if the test was posi- uniform censoring date (December 31, 2006), tive. In Italy, a PSA value of 4.0 ng per milliliter with identical follow-up in the two study groups. or more was defined as positive, but men with a Causes of death were evaluated in a blinded fash- PSA value of 2.5 to 3.9 ng per milliliter also un- ion and according to a standard algorithm9 or, derwent ancillary tests (digital rectal examination after validation, on the basis of official causes of and transrectal ultrasonography). death. The causes were classified by the indepen- In the Dutch and Belgian centers, up to Febru- dent committees as definite prostate cancer, causes ary 1997, a combination of digital rectal exami- related to screening, probable or possible prostate nation, transrectal ultrasonography, and PSA test- cancer, and other intercurrent causes (with or with- ing (with a cutoff value of 4.0 ng per milliliter) out prostate cancer as a contributory factor). Deci-1322 n engl j med 360;13 nejm.org march 26, 2009 The New England Journal of Medicine Downloaded from nejm.org on October 8, 2011. For personal use only. No other uses without permission. Copyright © 2009 Massachusetts Medical Society. All rights reserved.
  • 4. European Trial of Prostate-Cancer Screeningsion points that were used for determining the had the same underlying risk and that screeningcause of death have been described previously.9 in the control group was as effective as that inFor this analysis, we have combined the catego- the screening group.ries of definite and probable prostate cancer andthe category of causes related to screening. R e sult sOther Analyses SubjectsAspects of quality of life were evaluated in sev- Figure 1 shows trial enrollment, study-group as-eral study centers. A complete evaluation of all signments, and follow-up of all subjects and of thethe steps of screening was conducted in the core age group. A total of 162,387 men in the coreNetherlands (data not shown).15-21 age group underwent randomization; of these men, 72,952 were assigned to the screening group andStatistical Analysis 89,435 to the control group. A total of 62 men in theThe statistical analysis was based on the core age screening group and 82 men in the control groupgroup (including men between the ages of 55 and died between identification and randomization.69 years at randomization) and on the intention- Table 1 summarizes the characteristics of theto-screen principle. Overall mortality was studied subjects according to the center and the resultsto evaluate the correctness of randomization. of screening. The mean age at randomization wasPoisson regression analysis was used to estimate 60.8 years (range, 59.6 to 63.0), with little varia-the ratio of mortality in the intervention group to tion among the seven countries. In total, 82.2% ofmortality in the control group, stratified accord- the men in the screening group were screened ating to study center and age group at randomiza- least once. Compliance was higher in study cen-tion. The Nelsen–Aalen method was used for the ters that obtained consent before randomizationcalculation of cumulative hazard.22 All P values (88 to 100%) than in those in which subjects un-are two-sided. Interim analyses were conducted derwent randomization before providing consentfor follow-up in 2002, 2004, and 2006, with an (62 to 68%) (for details concerning all age groups,alpha spending curve with a division of uneven see Table 1A in Supplementary Appendix 5).weights.23 A preliminary analysis included men During the trial, 126,462 PSA-based tests werewho had actually undergone screening in the first performed, an average of 2.1 per subject whoround (with adjustment for noncompliance). The underwent screening. Overall, 16.2% of all testsnumber that would need to be screened to pre- were positive, with a range of 11.1 to 22.3%vent one death from prostate cancer was calcu- among the centers. The average rate of compli-lated as the inverse of the absolute difference in ance with biopsy recommendations was 85.8%cumulative mortality from prostate cancer be- (range, 65.4 to 90.3). Of the men who underwenttween the two study groups. biopsy for an elevated PSA value, 13,308 (75.9%) The study had a power of 86% to show a sta- had a false positive result.tistically significant difference of 25% or more in We detected 5990 prostate cancers in theprostate-cancer mortality with a P value of 0.05 screening group and 4307 in the control group.among men who underwent screening, on the These numbers correspond to a cumulative inci-basis of follow-up through 2008.4 The sample- dence of 8.2% and 4.8%, respectively. The posi-size calculation, which was part of the power tive predictive value of a biopsy (the number ofcalculation, took into account noncompliance in cancers detected on screening divided by thethe screening group in each study center and the number of biopsies expressed as a percentage)use of PSA tests outside the protocol assignment was on average 24.1% (range, 18.6 to 29.6). Thein the control group (termed contamination of cumulative incidence of local prostate cancer wasthe control group). On the basis of an overall higher in the screening group than in the controllevel of compliance of 82% and 20% contamina- group (for details about tumor stage, grade distri-tion in the control group, a 25% reduction in the bution, and treatment, see Supplementary Appen-number of men who underwent screening would dixes 6 and 7). For example, the number of menbe equivalent to a 14% reduction in an intention- with positive results on a bone scan (or a PSAto-treat analysis. This assumes that men who value of more than 100 ng per milliliter in thosewere screened and those who were not screened without bone-scan results) was 0.23 per 1000 n engl j med 360;13 nejm.org march 26, 2009 1323 The New England Journal of Medicine Downloaded from nejm.org on October 8, 2011. For personal use only. No other uses without permission. Copyright © 2009 Massachusetts Medical Society. All rights reserved.
  • 5. Table 1. Numbers of Subjects and Results of Screening, According to Study Center.* 1324 Variable The Netherlands Belgium Sweden Finland Italy Spain Switzerland Total November 1993– June 1991– June 1991– January 1996– October 1996– February 1996– September 1998– June 1991– March 2000 December 2003 December 2003 January 1999 October 2000 June 1999 August 2003 December 2003 Total no. of subjects 34,833 8562 11,852 80,379 14,517 2197 9903 162,243 Screening group — no. (%) 17,443 (50.1) 4307 (50.3) 5,901 (49.9) 31,970 (39.8) 7,265 (50.0) 1056 (48.1) 4948 (50.0) 72,890 (44.9) Control group — no. (%) 17,390 (49.9) 4255 (49.7) 5,951 (50.1) 48,409 (60.2) 7,252 (50.0) 1141 (51.9) 4955 (50.0) 89,353 (55.1) Age at randomization — yr All subjects Mean 61.9 63.0 59.8 59.6 62.2 61.0 61.6 60.8 Median 61.7 63.0 59.7 58.7 61.8 60.4 61.1 60.1 Screening group Mean 61.9 63.0 59.8 59.6 62.2 60.5 61.6 60.9 The Median 61.7 63.0 59.7 58.7 61.7 59.7 61.0 60.3 Control group Mean 62.0 63.0 59.8 59.6 62.2 61.4 61.7 60.7 Median 61.7 63.1 59.7 58.7 61.9 61.1 61.2 59.9 First round of screening — no. (%) 16,502 (94.6) 3795 (88.1) 3,649 (61.8) 20,796 (65.0) 4,961 (68.3) 1056 (100) 4721 (95.4) 55,480 (76.1) n engl j med 360;13 Screening interval — yr 4 4–7 2 4 4 4 4 NA Screened at least once — no. (%) 16,502 (94.6) 3876 (90.0) 4,466 (75.7) 23,608 (73.8) 5,675 (78.1) 1056 (100) 4740 (95.8) 59,923 (82.2) No. of screening tests performed 34,526 6042 14,848 48,900 11,377 1846 8923 126,462 nejm.org Positive PSA tests — no. (%) 7,707 (22.3) 984 (16.3) 2,751 (18.5) 5,528 (11.3) 1,267 (11.1) 354 (19.2) 1846 (20.7) 20,437 (16.2) n e w e ng l a n d j o u r na l The New England Journal of Medicine Biopsies — no. (%) 6,929 (89.9) 728 (74.0) 2,382 (86.6) 4,991 (90.3) 828 (65.4) 263 (74.3) 1422 (77.0) 17,543 (85.8) of Prostate cancers Total detected in screening 1,736 (10.0) 363 (8.4) 697 (11.8) 2,493 (7.8) 280 (3.9) 68 (6.4) 353 (7.1) 5,990 (8.2) march 26, 2009 group — no. (%) Copyright © 2009 Massachusetts Medical Society. All rights reserved. Detected during screening 1,521 182 550 1,477 180 60 265 4,235 m e dic i n e — no. Detected outside of 215 181 147 1,016 100 8 88 1,755 screening protocol — no. Positive predictive value of 22.0 25.0 23.1 29.6 21.7 22.8 18.6 24.1Downloaded from nejm.org on October 8, 2011. For personal use only. No other uses without permission. screening — %† Total detected in control group 685 (3.9) 252 (5.9) 421 (7.1) 2,632 (5.4) 133 (1.8) 24 (2.1) 160 (3.2) 4,307 (4.8) — no. (%) * The results are for the predefined core age group for this study, which included men between the ages of 55 and 69 years. The dates that are listed for each country are the periods in which subjects underwent randomization. NA denotes not applicable, and PSA prostate-specific antigen. † The positive predictive value of biopsy was calculated as the number of screen-detected cancers divided by the number of biopsies.
  • 6. European Trial of Prostate-Cancer Screeningperson-years in the screening group, as comparedwith 0.39 per 1000 person-years in the control 0.020 Nelson–Aalen Cumulative Hazardgroup, a 41% reduction in the screening group(P<0.001). The proportions of men who had a 0.015Gleason score of 6 or less were 72.2% in the Control groupscreening group and 54.8% in the control group, 0.010and the proportions with a Gleason score of 7 ormore were 27.8% in the screening group and 0.00545.2% in the control group. Screening group 0.000Prostate-Cancer Mortality 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14As of December 31, 2006, with average and medi- Years since Randomizationan follow-up times of 8.8 and 9.0 years in the No. at Riskscreening and control groups, respectively, there Screening group 65,078 58,902 20,288 Control group 80,101 73,534 23,758were 214 prostate-cancer deaths in the screeninggroup and 326 in the control group in the core Figure 2. Cumulative Risk of Death from Prostate Cancer.age group. Deaths that were associated with ICM AUTHOR: Schroeder RETAKE 1st As of December 31, 2006, with an average follow-up time of 8.8 years, there FIGURE: 2 of 2 2ndprostate-cancer–related interventions were cate- REG F were 214 prostate-cancer deaths in the screening group and 326 in the con- 3rdgorized as deaths from prostate cancer. The un- CASE that were associated with interventions were categorized trol group. Deaths Revised EMail Line 4-C SIZEadjusted rate ratio for death from prostate cancer as being due to prostate cancer. The adjusted rate ratio for death from prostate ARTIST: ts H/T H/T 22p3in the screening group was 0.80 (95% confidence cancer in the screening group was 0.80 (95% CI, 0.65 to 0.98; P = 0.04). The Enon Combo Nelsen–Aalen method was used for the calculation of cumulative hazard.interval [CI], 0.67 to 0.95; P = 0.01); after adjustment AUTHOR, PLEASE NOTE: Figure has been redrawn and type has been reset.for sequential testing with alpha spending due to Please check carefully.two previous interim analyses (based on Poissonregression analysis), the rate ratio was 0.80 (95% need to beJOB: 36013 (48) remained unchanged in 03-26-09 treated ISSUE:CI, 0.65 to 0.98; P = 0.04). The rates of death in the per-protocol analysis because the same num-the two study groups began to diverge after 7 to 8 ber of deaths were prevented and the same num-years and continued to diverge further over time ber of additional cases were diagnosed in men(Fig. 2). Overall mortality results at 30 days are who actually underwent screening.summarized in Supplementary Appendix 8. In the intention-to-screen analysis, the abso- Effect of Age on Mortalitylute difference between the screening group and In an exploratory analysis of mortality accordingthe control group was 0.71 prostate-cancer death to age group, there was no evidence of heteroge-per 1000 men. This means that in order to pre- neity among age groups (Table 2). Among menvent one prostate-cancer death, the number of between the ages of 50 and 54 years at baseline,men who would need to be screened would be the number of events was small, with no obvious1410 (95% CI, 1142 to 1721), with an average of screening effect.1.7 screening visits per subject during a 9-yearperiod. The additional prostate cancers diagnosed Heterogeneity of Rate Ratiosby screening resulted in an increase in cumulative In an exploratory analysis of heterogeneity accord-incidence of 34 per 1000 men, as compared with ing to study center (which was carried out in ac-the control group. In other words, 48 additional cordance with the monitoring plan6), the decreasesubjects (1410 ÷ 1000 × 34) would need to be treat- in the rate of death from prostate cancer in theed to prevent one death from prostate cancer. screening group could not be attributed to any In an analysis of men who were actually single center, as evidenced by rate ratios rangingscreened during the first round (which was ad- between 0.74 and 0.84 after the exclusion of eachjusted for noncompliance), the rate ratio for center, one at a time. There was no significantprostate-cancer death after 9 years was 0.73 difference in overall mortality (Table 3).(95% CI, 0.56 to 0.90), which meant that 1068men would need to be screened and 48 would Adverse Eventsneed to be treated to prevent one death from No deaths were reported as a direct complicationprostate cancer. The number of men who would (e.g., septicemia or bleeding) associated with a n engl j med 360;13 nejm.org march 26, 2009 1325 The New England Journal of Medicine Downloaded from nejm.org on October 8, 2011. For personal use only. No other uses without permission. Copyright © 2009 Massachusetts Medical Society. All rights reserved.
  • 7. The n e w e ng l a n d j o u r na l of m e dic i n e Table 2. Death from Prostate Cancer, According to the Age at Randomization.* Age at Randomization Screening Group Control Group Rate Ratio (95% CI)† Person-Yr (Death Person-Yr (Death No. of Rate per 1000 No. of Rate per 1000 Deaths Person-Yr) Deaths Person-Yr) All subjects 261 737,397 (0.35) 363 878,547 (0.41) 0.85 (0.73–1.00) Age group 50–54 yr 6 55,241 (0.11) 4 53,734 (0.07) 1.47 (0.41–5.19) 55–59 yr 60 316,389 (0.19) 102 402,062 (0.25) 0.73 (0.53–1.00) 60–64 yr 76 191,542 (0.40) 95 221,113 (0.43) 0.94 (0.69–1.27) 65–69 yr 78 135,470 (0.58) 129 162,410 (0.79) 0.74 (0.56–0.99) 70–74 yr 41 38,755 (1.06) 33 39,228 (0.84) 1.26 (0.80–1.99) * The result of the chi-square test for heterogeneity among subjects in the core age group (55 to 69 years) was 2.44 (P = 0.49). † Rate ratios were calculated with the use of Poisson regression and compare the rate of death from prostate cancer in the screening group with the rate in the control group. biopsy procedure. Complications associated with sults were based on a combined analysis of data screening procedures (including prostate biopsy) from centers sharing a common core protocol, have been reported previously.24,25 which defined the minimal criteria for inclusion and the scope of the primary analysis but allowed Discussion wider age ranges or shorter screening intervals. Because of various recruitment approaches, the In an intention-to-screen analysis of data from estimate of a 20% reduction in prostate-cancer seven European centers, PSA screening was as- mortality does not represent the effect of a screen- sociated with a significant absolute reduction of ing program at the population level or the effect 0.71 prostate-cancer death per 1000 men after an on individual subjects but instead represents a average follow-up of 8.8 years (median, 9.0). This mixture of such estimates. Despite some varia- finding corresponds to a relative reduction of tion in screening procedures, the results from each 20% in the rate of death from prostate cancer center were compatible with the main result: among men between the ages of 55 and 69 years at a lowering of the death rate from prostate cancer study entry, given an average screening interval of associated with screening. 4 years and a compliance rate of 82% of those The screening interval of 4 years was chosen who accepted the offer of screening (rate ratio, on the basis of the mean lead time of 5 to 10 0.80; adjusted P = 0.04). To prevent one prostate- years in PSA-based screening.28,29 However, the cancer death, 1410 men (or 1068 men who actually lead time of aggressive cancers, which may be underwent screening) would have to be screened, the most important target of screening, is likely and an additional 48 men would have to be treated. to be much shorter. The high number of men who would need to be The benefit of screening was restricted to the treated could be improved by avoiding the diag- core age group of subjects who were between the nosis and treatment of indolent cancers during ages of 55 and 69 years at the time of random- screening or by improving treatment in the re- ization. The results that were seen in other age maining men with cancer. The number needed to groups are preliminary and inconclusive. Our find- screen in our study is similar to that in studies of ings are early results of the trial, and continued mammographic screening for breast cancer and follow-up will provide further information. Ad- fecal occult-blood testing for colorectal cancer.26,27 justment for noncompliance resulted in a greater Our analysis shows that the results were gen- effect among men who actually underwent screen- erally similar in all participating study centers ing, and after adjustment for both noncompli- considered individually (Table 3). The trial was ance and contamination, the effect of screening not powered to evaluate mortality differences in the intention-to-screen analysis is likely to be between centers or for age subgroups. The re- further enhanced.1326 n engl j med 360;13 nejm.org march 26, 2009 The New England Journal of Medicine Downloaded from nejm.org on October 8, 2011. For personal use only. No other uses without permission. Copyright © 2009 Massachusetts Medical Society. All rights reserved.
  • 8. European Trial of Prostate-Cancer Screening Table 3. Rate Ratios for Death from Any Cause and Death from Prostate Cancer, with Exclusions According to Location of Study Center.* Variable Rate Ratio (95% CI) P Value† All deaths from any cause 0.99 (0.97–1.02) 0.50 All deaths from prostate cancer 0.80 (0.67–0.95) 0.01 Excluding the Netherlands 0.81 (0.67–0.99) 0.04 Excluding Finland 0.74 (0.58–0.94) 0.01 Excluding Sweden 0.84 (0.70–1.01) 0.06 Excluding Belgium 0.79 (0.66–0.94) 0.01 Excluding Spain 0.79 (0.67–0.94) 0.01 Excluding Italy 0.79 (0.66–0.94) 0.01 Excluding Switzerland 0.80 (0.68–0.96) 0.02* Rate ratios, which were calculated with the use of Poisson regression, compare the rate of death from prostate cancer in the screening group with the rate in the control group. The calculations were restricted to men in the core age group (55 to 69 years).† P values have not been corrected for multiple testing. The rate of overdiagnosis of prostate cancer ulation coverage, overdiagnosis, overtreatment,(defined as the diagnosis in men who would not quality of life, cost, and cost-effectiveness. Thehave clinical symptoms during their lifetime) has ratio of benefits to risks that is achievable withbeen estimated to be as high as 50% in the more frequent screening or a lower PSA thresh-screening group.30 Consistent estimates of over- old than we used remains unknown. Further A video roundtablediagnosis (a third of cancers detected on screen- analyses are needed to determine the optimal and comments on the value ofing) have also been obtained by identifying po- screening interval in consideration of the PSA PSA screeningtentially indolent prostate cancers on the basis value at the first screening and of previously are available atof clinical and pathological characteristics.31-33 negative results on biopsy.35-38 NEJM.orgOverdiagnosis and overtreatment are probably Supported by grants from Europe Against Cancer and thethe most important adverse effects of prostate- fifth and sixth framework program of the European Union, bycancer screening and are vastly more common grants from agencies or health authorities in the participatingthan in screening for breast, colorectal, or cervi- countries, and by unconditional grants from Beckman Coulter. The studies in each national center were funded by numerouscal cancer.34 local grants (see Supplementary Appendix 2). Although the results of our trial indicate a Dr. Hugosson reports receiving lecture fees from GlaxoSmith-reduction in prostate-cancer mortality associated Kline; Dr. Tammela, receiving consulting or lecture fees from GlaxoSmithKline, Pfizer, AstraZeneca, Leiras, and Novartis; andwith PSA screening, the introduction of popula- Dr. Lilja, holding a patent for an assay for free PSA. No othertion-based screening must take into account pop- potential conflict of interest relevant to this article was reported. AppendixThe authors’ affiliations are as follows: the Departments of Urology (F.H.S., M.J.R., C.H.B.), Pathology (T.K.), Clinical Chemistry(B.G.B.), and Public Health (H.J.K.), Erasmus Medical Center, Rotterdam, the Netherlands; the Department of Urology, SahlgrenskaUniversity Hospital, Göteborg, Sweden (J.H., G.A.); the Department of Urology, Tampere University Hospital (T.L.J.T.), and TampereSchool of Public Health, University of Tampere (A.A.) — both in Tampere, Finland; the Department of Diagnostic Medical Imaging(S.C.) and Unit of Epidemiology (M.Z.), Institute for Cancer Prevention, Florence, Italy; Provinciaal Instituut voor Hygiëne, Antwerp,Belgium (V.N.); the Department of Urology, Kantonsspital Aarau, Aarau, Switzerland (M.K., F.R.); the Department of Urology, HospitalUniversitario de Getafe, Madrid (M.L., A.B.); the Department of Laboratory Medicine, Lund University, Malmö University Hospital,Malmö, Sweden, and Memorial Sloan-Kettering Cancer Center, New York (H.L.); Oncology Center Antwerp, Antwerp, Belgium (L.J.D.);Finnish Cancer Registry, Helsinki (L.M.); the Department of Urology, Centre Hospitalier Regional Universitaire, Lille, France (A.V.); theDepartment of Urology, Clinique de Beau Soleil, Montpellier, France (X.R.); and the Cancer Screening Evaluation Unit, Surrey, UnitedKingdom (S.M.M.). The members of the data and safety monitoring committee were as follows: P. Smith (chair), J. Adolfsson, and T. Hakulinen, whocarried out interim analyses of the data by relating the central messages on the progress of the study to the voting members; J. Cham-berlain and B. Collette, earlier committee members; F. Alexander, who served as the trial statistician until her retirement; and I. deBeaufort, who served as an advisor. Additional study members are listed in Supplementary Appendix 1. n engl j med 360;13 nejm.org march 26, 2009 1327 The New England Journal of Medicine Downloaded from nejm.org on October 8, 2011. For personal use only. No other uses without permission. Copyright © 2009 Massachusetts Medical Society. All rights reserved.
  • 9. The n e w e ng l a n d j o u r na l of m e dic i n e References 1. Stamey TA, Yang N, Hay AR, McNeal 15. Essink-Bot ML, de Koning HJ, Nijs gen and its complex with alpha 1-antichy- JE, Freiha FS, Redwine E. Prostate-specific HGT, Kirkels WJ, van der Maas PJ, Schröder motrypsin before diagnosis of prostate antigen as a serum marker for adenocar- FH. Short-term effects of population- cancer. Lancet 1994;344:1594-8. cinoma of the prostate. N Engl J Med based screening for prostate cancer on 29. Gann PH, Hennekens CH, Stampfer 1987;317:909-16. health-related quality of life. J Natl Can- MJ. A prospective evaluation of plasma 2. Catalona WJ, Smith DS, Ratliff TL, et al. cer Inst 1998;90:925-31. prostate-specific antigen for detection of Measurement of prostate-specific antigen 16. Madalinska JB, Essink-Bot ML, de prostate cancer. JAMA 1995;273:289-94. in serum as a screening test for prostate Koning HJ, Kirkels WJ, van der Maas PJ, 30. Draisma G, Boer R, Otto SJ, et al. Lead cancer. N Engl J Med 1991;324:1156-61. Schröder FH. Health-related quality-of- times and overdetection due to prostate- [Erratum, N Engl J Med 1991;325:1324.] life effects of radical prostatectomy and specific antigen screening: estimates from 3. Denis LJ, Murphy GP, Schröder FH. primary radiotherapy for screen-detected the European Randomized Study of Screen- Report of the consensus workshop on of clinically diagnosed localized prostate ing for Prostate Cancer. J Natl Cancer Inst screening and global strategy for prostate cancer. J Clin Oncol 2001;19:1619-28. 2003;95:868-78. cancer. Cancer 1995;75:1187-207. 17. Idem. Health-related quality of life in 31. Steyerberg EW, Roobol MJ, Kattan 4. de Koning HJ, Liem MK, Baan CA, patients with screen-detected versus clini- MW, van der Kwast TH, de Koning HJ, Boer R, Schröder FH, Alexander FE. Pros- cally diagnosed prostate cancer preceding Schröder FH. Prediction of indolent pros- tate cancer mortality reduction by screen- primary treatment. Prostate 2001;46:87-97. tate cancer: validation and updating of a ing: power and time frame with complete 18. Korfage IJ, Essink-Bot ML, Madalin- prognostic nomogram. J Urol 2007;177: enrollment in the European Randomised ska JB, Kirkels WJ, Litwin MS, de Koning 107-12. Screening for Prostate Cancer (ERSPC) HJ. Measuring disease specific quality of 32. Khatami A, Aus G, Damber JE, Lilja trial. Int J Cancer 2002;98:268-73. life in localized prostate cancer: the Dutch H, Lodding P, Hugosson J. PSA doubling 5. Andriole GL, Crawford ED, Grubb RL experience. Qual Life Res 2003;12:459-64. time predicts the outcome after active III, et al. Mortality results from a random- 19. Essink-Bot ML, Korfage IJ, de Koning surveillance in screening-detected prostate ized prostate-cancer screening trial. N Engl HJ. Including the quality-of-life aspects in cancer: results from the European Ran- J Med 2009:360;1310-9. the evaluation of prostate cancer screen- domized Study of Screening for Prostate 6. De Koning HJ, Hakulinen T, Moss ing: expert opinions revisited? BJU Int Cancer, Sweden section. Int J Cancer 2007; SM, Adolfsson J, Smith PH, Alexander FE. 2003;92:Suppl 2:101-5. 120:170-4. Monitoring the ERSPC trial. BJU Int 2003; 20. Korfage IJ, de Koning HJ, Roobol MJ, 33. Roemeling S, Roobol MJ, Kattan MW, 92:Suppl 2:112-4. Schröder FH, Essink-Bot ML. Prostate can- van der Kwast TH, Steyerberg EW, 7. Roobol MJ, Schröder FH. European cer diagnosis: the impact on patients’ men- Schröder FH. Nomogram use for the pre- Randomized study of Screening for Pros- tal health. Eur J Cancer 2006;42:165-70. diction of indolent prostate cancer: impact tate Cancer (ERSPC): rationale, structure 21. Korfage IJ, de Koning HJ, Habbema JD, on screen-detected populations. Cancer and preliminary results. BJU Int 2003;92: Schröder FH, Essink-Bot ML. Side-effects 2007;110:2218-21. Suppl 2:1-122. of treatment for localized prostate cancer: 34. Hakama M, Auvinen A. Cancer screen- 8. Smith PH. The Data Monitoring Com- are they valued differently by patients and ing. In: Heggenhougen K, Quah SR, eds. mittee — bridging the gap between urol- healthy controls? BJU Int 2007;99:801-6. International encyclopedia of public health. ogy and public health epidemiology. BJU 22. Aalen OO. Nonparametric inference San Diego, CA: Academic Press, 2008:464- Int 2003;92:Suppl 2:55-6. for a family of counting processes. Ann 80. 9. De Koning HJ, Blom J, Merkelbach Stat 1978;6:701-26. 35. Roobol MJ, Grenabo A, Schröder FH, JW, et al. Determining the cause of death 23. O’Brien PC, Fleming TR. A multiple Hugosson J. Interval cancers in prostate in randomized screening trial(s) for pros- testing procedure for clinical trials. Bio- cancer screening: comparing 2- and 4-year tate cancer. BJU Int 2003;92:Suppl 2:71-8. metrics 1979;35:549-56. screening intervals in the European Ran- 10. Laffin RJ, Chan DW, Tanasijevic MJ, 24. Raaijmakers R, Kirkels WJ, Roobol domized Study of Screening for Prostate et al. Hybritech total and free prostate MJ, Wildhagen MF, Schröder FH. Compli- Cancer, Gothenburg and Rotterdam. J Natl specific antigen assays developed for the cation rates and risk factors of 5802 trans- Cancer Inst 2007;99:1296-303. Beckman Coulter access automated chemi- rectal ultrasound-guided sextant biopsies 36. Roobol MJ, Schröder FH, Kranse R. luminescent immunoassay system: a multi- of the prostate within a population-based A comparison of first and repeat (four center evaluation of analytical perfor- screening program. Urology 2002;60:826- years later) prostate cancer screening in a mance. Clin Chem 2001;47:129-32. 30. randomized cohort of asymptomatic men 11. Schröder FH, Van der Maas PJ, Beem- 25. Mäkinen T, Auvinen A, Hakama M, aged 55-75 years using a biopsy indication sterboer PMM, et al. Evaluation of the Stenman UH, Tammela TL. Acceptability of 3.0 ng/ml (results of ERSPC, Rotter- digital rectal examination as a screening and complications of prostate biopsy in dam). Prostate 2006;66:604-12. test for prostate cancer. J Natl Cancer Inst population-based PSA screening versus 37. Roobol MJ, Roobol DW, Schröder FH. 1998;90:1817-23. routine clinical practice: a prospective, Is additional testing necessary in men 12. Schröder FH, Roobol-Bouts M, Vis AN, controlled study. Urology 2002;60:846-50. with prostate-specific antigen levels of van der Kwast T, Kranse R. Prostate-spe- 26. IARC handbooks of cancer preven- 1.0 ng/mL or less in a population based cific antigen-based early detection of pros- tion. Vol. 7. Breast cancer screening. Lyon, screening setting? (ERSPC, section Rotter- tate cancer — validation of screening with- France: International Agency for Research dam.) Urology 2005;65:343-6. out rectal examination. Urology 2001;57: on Cancer, 2002. 38. Aus G, Damber JE, Khatami A, Lilja 83-90. 27. Hewitson P, Glasziou P, Watson E, H, Stranne J, Hugosson J. Individualized 13. Stamey TA. Making the most out of Towler B, Irwig L. Cochrane systematic screening interval for prostate cancer six systematic sextant biopsies. Urology review of colorectal cancer screening us- based on prostate-specific antigen level: 1995;45:2-12. ing the fecal occult blood test (Hemoccult): results of a prospective, randomized, 14. Wolters T, Roobol MJ, Steyerberg EW, an update. Am J Gastroenterol 2008;103: population-based study. Arch Intern Med et al. No evidence for a treatment bias in 1541-9. 2005;165:1857-61. the European Randomized Study of Screen- 28. Stenman UH, Hakama M, Knekt P, Copyright © 2009 Massachusetts Medical Society. ing for Prostate Cancer. Int J Cancer (in Aromaa A, Teppo L, Leinonen J. Serum press). concentrations of prostate specific anti-1328 n engl j med 360;13 nejm.org march 26, 2009 The New England Journal of Medicine Downloaded from nejm.org on October 8, 2011. For personal use only. No other uses without permission. Copyright © 2009 Massachusetts Medical Society. All rights reserved.